Nanyun Peng
Associate Professor
Researches controllable and creative language generation, multimodal language models and evaluation of language model reasoning.
Nanyun Peng's lab at UCLA works on large language models, reasoning & agents and vision-language & multimodal. 136 papers since 2023, 87 at top venues, mostly EMNLP and ACL. 9 PhD students. 35 alumni, most often to Google and Amazon. Applications close .
At a glance
- papers, 2023–2026
- 35/ year
- at T1 venues
- 64%
- PhD students
- 9
- MS students
- 0
- alumni listed
- 35
Papers by yearfrom the lab homepage, 2023–2026
- 30 papers2026 · T1 15 · collecting
- 34 papers2025 · T1 25
- 42 papers2024 · T1 25
- 30 papers2023 · T1 22
T1 venuesOtherVenues: EMNLP, ACL, Findings of ACL, NeurIPS, ICML
Research focuspaper titles that use each topic's words
- Large language models60 papers
- Vision-language & multimodal46 papers
- language34 papers
- Reasoning & agents22 papers
- model reasoning0 papers
Paper list
From the lab homepage.
202630 papers
- ICLR (T1)ARES: Multimodal Adaptive Reasoning via Difficulty-Aware Token-Level Entropy Shaping
- Findings of ACLBeyond Facts- Benchmarking Distributional Reading Comprehension in Large Language Models
- Findings of ACLBRIEF-Pro: Universal Context Compression with Short-to-Long Synthesis for Fast and Accurate Multi-Hop Reasoning
- ACL (T1)Decoupling Task-Solving and Output Formatting in LLM Generation
- NeurIPS (T1)Does Inference-Time Reasoning Really Improve Video Understanding?
- JUDGeDoes Reasoning Improve Psychological Depth in Large Language Models? Depends on Who's Judging
- ICLR (T1)Energy-Regularized Sequential Model Editing on Hyperspheres
- EMNLP (T1)Extracting Small Translation Specialists from LLMs by Aggressively Pruning Experts
- COLMFronTalk: Benchmarking Front-End Development as Conversational Code Generation with Multi-Modal Feedback
- NonAR-LM WorkshopInsNeXt: Scalable Pre-training of Efficient Insertion-based Diffusion Language Models
- NeurIPS (T1)LEAP: Supercharging LLMs for Formal Mathematics with Agentic Frameworks
- ICML (T1)Learning Structured Reasoning via Tractable Trajectory Control
202534 papers
- NeurIPS (T1)3DLLM-Mem: Long-Term Spatial-Temporal Memory for Embodied 3D Large Language Model
- NeurIPS (T1)AI Debate Aids Assessment of Controversial Claims
- NAACL-FindingsBRIEF: Bridging Retrieval and Inference for Multi-hop Reasoning via Compression
- EMNLP (T1)CaKE: Circuit-aware Editing Enables Generalizable Knowledge Learners
- ACL (T1)Collapse of Dense Retrievers: Short, Early, and Literal Biases Outranking Factual Evidence
- Findings of ACLComparing Bad Apples to Good Oranges: Aligning Large Language Models via Joint Preference Optimization
- ICML (T1)Contrastive Visual Data Augmentation
- EMNLP (T1)DiCoRe: Enhancing Zero-shot Event Detection via Divergent-Convergent LLM Reasoning
- Findings of ACLDRS: Deep Question Reformulation With Structured Output
- NAACL-FindingsEvaluating Cultural and Social Awareness of LLM Web Agents
- EMNLP-FindingsFLAMES: Improving LLM Math Reasoning via a Fine-Grained Analysis of the Data Synthesis Pipeline
- TMLR (T1)FlexEControl: Flexible and Efficient Multimodal Control for Text-to-Image Generation
202442 papers
- NeurIPS (T1)Adaptable Logical Control for Large Language Models
- NAACL (T1)AMRFact: Enhancing Summarization Factuality Evaluation with AMR-Driven Negative Samples Generation
- EMNLP (T1)Are Large Language Models Capable of Generating Human-Level Narratives?
- ACL-FindingsArgument-Aware Approach To Event Linking (opens in a new tab)
- EMNLP WorkshopARMADA: Attribute-Based Multimodal Data Augmentation
- Findings of ACLCaLM: Contrasting Large and Small Language Models to Verify Grounded Generation
- NAACL (T1)Contextual Label Projection for Cross-Lingual Structured Prediction
- ICML (T1)ConTextual: Evaluating Context-Sensitive Text-Rich Visual Reasoning in Large Multimodal Models
- Proceedings of The Thirty- Annual Conference on Neural Information Processing Systems (NeurIPS), Datasets and Benchmarks TrackDACO: Towards Application-Driven and Comprehensive Data Analysis via Code Generation
- Findings of ACLDetecting Machine-Generated Long-Form Content with Latent-Space Variables
- ICML (T1)DiNADO: Norm-Disentangled Neurally-Decomposed Oracles for Controlling Language Models
- EMNLP (T1)Do LLMs Plan Like Human Writers? Comparing Journalist Coverage of Press Releases with LLMs
202330 papers
- EMNLP-Findings“Kelly is a Warm Person, Joseph is a Role Model”: Gender Biases in LLM-Generated Reference Letters
- ACL (T1)ACCENT: An Automatic Event Commonsense Evaluation Metric for Open-Domain Dialogue Systems
- EMNLP (T1)ACQUIRED: A Dataset for Answering Counterfactual Questions In Real-Life Videos
- EMNLP (T1)Active Instruction Tuning: Improving Cross-Task Generalization by Training on Prompt Sensitive Tasks
- ACL (T1)AMPERE: AMR-Aware Prefix for Generation-Based Event Argument Extraction Model
- ACL (T1)Are Fairy Tales Fair? Analyzing Gender Bias in Temporal Narrative Event Chains of Children's Fairy Tales
- EMNLP-FindingsAre Personalized Stochastic Parrots More Dangerous? Evaluating Persona Biases in Dialogue Systems
- Findings of ACLCode-Switching Text Synthesis in Unseen Language Pairs
- NeurIPS (T1)DesCo: Learning Object Recognition with Rich Language Descriptions
- ACL (T1)DICE: Data-Efficient Clinical Event Extraction with Generative Models
- Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL), shortDo Models Really Learn to Follow Instructions? An Empirical Study of Instruction Tuning
- ACL (T1)DOC: Improving Long Story Coherence With Detailed Outline Control
Where alumni went35 listed, 58% to industry
Industry 11Academia 8
Most often to
- Amazon
- Apple
Also to
- Carnegie Mellon University
- NYU
- Stanford University
Apply to UCLAFall 2027
Computer Science PhD
- Round
- Single round (Dec 15)
- Opens
- —
- GRE
- optional
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- USMIT
Paul Liang
21 T1 - USUC Berkeley
Dan Klein
- Multilingual, translation & linguistics
- USUCLA
Kai-Wei Chang
- Large language models
- USNYU
Kyunghyun Cho
- Large language models
- GBEdinburgh
Alexandra Birch
- Multilingual, translation & linguistics
- SGNTU
Soujanya Poria
- Large language models
Your notes
Save this lab to keep notes.
Data
Checked against the lab homepage.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
UCLAcloses